cavekit is a Claude Code plugin that organizes software changes through a compact, spec-driven workflow stored in a SPEC.md file. It is for developers who want to plan, research, review, and build changes while preserving project context across sessions; the catalogue entries are its skills, commands, plugin configuration, and workflow rules.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/JuliusBrussee/cavekitWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/juliusbrussee/cavekit/build)<a href="https://agentmods.dev/commands/juliusbrussee/cavekit/build"><img src="https://agentmods.dev/badge/commands/juliusbrussee/cavekit/build/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/juliusbrussee/cavekit/build"><img src="https://agentmods.dev/badge/commands/juliusbrussee/cavekit/build.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00019 | $0.00144 |
| Opus 5.5 | $0.00008 | $0.00058 |
| Sonnet 5.5 | $0.00004 | $0.00029 |
| Haiku 4.5 | $0.00002 | $0.00014 |
Grade A, and why
build scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
Invoke the build skill (skills/build/SKILL.md). Treat $ARGUMENTS as the
target (§T.n, --next, or --all). Plan in native plan mode, name the exact
test that proves each §V touched (verification contract), read §R for external
facts, execute, and auto-invoke backprop on failure. High blast radius? Suggest
/ck:review first. Build only flips §T status; other spec edits route through spec.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 11 lines · 19 tokens per session scan A 8a10a3a8d9d9
build is a command published in the GitHub repository JuliusBrussee/cavekit (1,154 stars, last pushed 2mo ago), licensed MIT. It adds 19 tokens to every session and 144 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other commands, from other repositories
cc-test
(Conductor) Analyze coverage gaps and write missing tests.
start
Launch the Consort workflow (spec-first design, test-driven build on paired Lakebase branches). In a scaffolded project, takes stock and resumes the /plan -> /design -> /build -> /deploy loop; elsewhere, guides you through creating a project, then resumes.
design
Drives a feature from idea to spec to architect review to test list. This wraps the canonical consort design phases as a single one-shot you can invoke from Claude Code in a Lakebase-paired project.
build
Drives a designed feature through TDD cycles to ready-for-review by delegating to the deterministic orchestrator driver. It is the second phase of the per-feature loop: /design -> /build -> /deploy, which a sprint of features runs after /plan authored their requests. When the stories are built + accepted, run /deploy…
test-tdd
Run when user calls /test-tdd. Scans modified files, locates their corresponding unit/integration test suites, and runs them.
complete
Complete a partially implemented feature by filling gaps and ensuring production readiness.